Computational single-cell methods for predicting cancer risk.

Area of Science:

• Computational biology
• Systems biology
• Cancer research

Background:

• Cancer risk prediction is a significant challenge despite biotechnological advances.
• Improved prediction is crucial for enhancing prevention, early detection, and survival rates.
• Current computational and experimental methods face limitations.

Purpose of the Study:

• To summarize emerging computational challenges and advances in cancer risk prediction.
• To focus on computational strategies utilizing single-cell data for cancer risk assessment.
• To introduce novel bottom-up network modeling approaches for estimating cancer stemness and dedifferentiation.

Main Methods:

• Utilizing single-cell omics data (scRNA-seq, snRNA-seq).
• Employing bottom-up network modeling from a systems-biological perspective.
• Describing two methods: diffusion network entropy (tissue/lineage-independent) and transcription factor regulons (tissue/lineage-specific).

Main Results:

• The developed computational tools successfully delineate the heterogeneous inter-cellular cancer-risk landscape.
• These methods can identify cells with a higher likelihood of developing into cancerous cells.
• Application to pre-invasive cancer stages demonstrated efficacy in risk assessment.

Conclusions:

• Bottom-up systems biological modeling of single-cell omic data represents a novel computational paradigm.
• This approach promises to advance the development of preventive and early cancer detection strategies.
• It facilitates more accurate cancer-risk prediction at the single-cell level.